Research On The 3D Audio Recording And Coding Technology Based On Distance Perception

With the rapid development of 3D movie,3D audio and video technology becomes the focus of academic.3D movie is developing from cinemas to home theaters and mobile terminals.3D audio and video technology for home theaters and mobile terminals has become a research topic.3D audio technology,as an important part of 3D audio and video technology,can effectively improve the immersive experience of 3D movie.Compared with the traditional audio,the number of 3D audio objects and the precision of 3D audio location in three-dimensional space are greatly increasing,which lead to a significant increase of the bit rate for high quality 3D audio.However,traditional audio coding technology cannot efficiently compress the signals and spatial parameters of 3D audio,which lead to the mobile network bandwidth is difficult to transmit the coding stream of high quality 3D audio,and limit the application and development of 3D audio technology.Focusing on the existing problems of the 3D audio coding technology,we measured and analyzed the distance perceptual sensitivity of the human's ear in the whole 3D space,and model of hearing perception sensitivity is established by combining with the existing perception sensitivity of azimuth and elevation.Then the model of perception sensitivity was used to guide the undistorted coding and clustering coding of 3D audio,which has obtained certain research results.The main research works and innovations of this paper are as follows:(1)The research of distance perception mechanism in three-dimensional acoustic fieldAt present,the research to the features of the spatial perception is mainly carried out on azimuth and elevation,the research of distance perception sensitivity are only in some azimuths.The sensitivity of distance perception in the whole three-dimensional space isn't modeled and analysed.Therefore,we couldn't explore the perception mechanism of three-dimensional acoustic field,which is insufficient to guide the recording,coding and reconstruction of 3D audio.In order to solve this problem,this paper,based on the perception sensitivity research results of azimuth and elevation,carried out the capturing and analyzing to the sensitivity of distance perception in the three dimensional space,and the sensitivity model of distance perception in three-dimensional space is established.Then,the sensitivity model of hearing perception in three-dimensional space is established by combining the sensitivity model of distance perception with the existing perception sensitivity of azimuth and elevation.The sensitivity model of hearing perception provides a theoretical basis for the recording,coding and reconstruction of 3D audio.Specific research contents as follow.Design the distance subjective perception listening experiment in three-dimensional space.Research the distance sensitivity of human ear in the three dimensional space.Establish the sensitivity model of distance perception in three-dimensional space.Establish the sensitivity model of hearing perception in three-dimensional space.(2)Undistorted coding of 3D audio based on distance perceptionThe traditional coding methods of 3D audio mainly consider the compression of parameters in the azimuth and elevation,and did not consider the extracting and compressing of distance information.The bit rate of 3D audio is limited by the bandwidth of wireless transmission channel,which result in distortion of distance perception in 3D audio and the waste of quantitative resources.Based on the sensitivity model of hearing perception,this paper proposed a method to calculate the amount of information for spatial parameters perception coding with undistorted,revealing the perception compression limits of 3D audio spatial parameters,designing the codebook of spatial parameters based on the sensitivity model of hearing perception,providing theoretical guidance for the undistorted perception compression coding method of 3D audio spatial parameters.Specific research contents as follow.Research the method to calculate the amount of information for spatial parameters perception coding with undistorted.Design the codebook of spatial parameters based on the sensitivity model of hearing perception.Research the undistorted coding method of spatial parameters based on perception for 3D audio.(3)The clustering coding of 3D audio based on perceptionThe compression of spatial parameters for 3D audio mainly considers the information redundancy between frames,using inter-frame difference coding to remove information redundancy between frames,but the information redundancy of spatial parameters within the frames are lack of consideration.With the improvement of space quantization precision and the increasing number of audio objects,the bit rate of spatial parameters are increasing,which result in the mobile network bandwidth has been unbearable for the existing spatial parameters coding.Inter-frame difference coding cannot meet the demand of 3D audio.Basing on the intra-frame spatial correlation of spatial parameters,the clustering coding of 3D audio based on perception is proposed,and the intra-frame compression bounds of 3D audio is revealed,which provides a theoretical guidance to 3D audio intra-frame compression.Specific research contents as follow.Capture the spatial parameters of 3D audio.Cluster the spatial parameter based on perception.Quantify and code the spatial parameters.